fix(debuginfo): disable overflow check for recursive non-enum types

[godzie44]

Commit b10edb4 introduce an overflow check when generating debuginfo for expanding recursive types. While this check works correctly for enums, it can incorrectly prune valid debug information for structures.

For example see #143241 (#143241 (comment)). Furthermore, for structures such check does not make sense, since structures with recursively expanding types simply will not compile (there is a hir_analysis_recursive_generic_parameter for that).

closes#143241

[rustbot]

r? @fee1-dead

rustbot has assigned @fee1-dead.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

[fee1-dead]

r? compiler

[fee1-dead]

cc @wesleywiser and @adwinwhite, this fixes regression #143241 most possibly caused by #138599

This should probably be tested

[tmiasko]

In principle, the original issue also applies to structs. For example:

use std::marker::PhantomData;pubstructAB<T:'static>{puba:PhantomData<T>,pubb:*constAB<(T,T)>,}fnmain(){
std::hint::black_box(AB::<()>{a:PhantomData,b: std::ptr::null()});}

[godzie44]

Sad, i miss this example before.

Btw, without the pub a: PhantomData<T> field, the hir_analysis_recursive_generic_parameter diagnostic is triggered and example are not complied. There may be some inconsistency here, according to hir_analysis_recursive_generic_parameter check?

[petrochenkov]

r? @wesleywiser as the reviewer of #138599

[rustbot]

wesleywiser is not on the review rotation at the moment.
They may take a while to respond.

[wesleywiser]

Can you add a codegen regression test that verifies we emit the appropriate amount of debuginfo for the minimal repro you found?

https://github.com/rust-lang/rust/blob/master/tests/codegen-llvm/debuginfo-inline-callsite-location.rs might be a helpful example.

[godzie44]

@rustbot ready

[wesleywiser]

Thanks @godzie44!

@bors r+

[bors]

📌 Commit 49eda8e has been approved by wesleywiser

It is now in the queue for this repository.